图像8邻域分形块编码  

Eight Neighbourhoods' Fractal Block Coding

作　　者：皮明红[1] 马俊 

机构地区：[1]华中理工大学,武汉430074

出　　处：《武汉纺织工学院学报》1997年第2期62-66,共5页

摘　　要：提出了8邻域分形块编码，其值域块的分形编码是基于对值域块最小均方差近似得来的。因而与基于搜索匹配的Jackquin分形块编码相比，它速度快，压缩比高。对512×512的“Lenna”图像实验结果表明:二级8邻域分形块编码每像素比特率可达到0．36bpp，峰值信噪比可达29．86dB，在PC486上，编码时间为64秒，解码时间为109秒。In this paper Eight Neighbourhoods Fractal Block Coding (ENFBC.) is presented.Fractal code of each range bloqk is based on a least.mean square error approximation in its eightneighbourhoods. Compared with Jackquin s fractal block coding, ENFBC's encoding time isinstantaneous and its compression ratio is ralatively high. Our conclusion indicates that ENFBCfor 512×512 'Lenna' image encodes to a compression ratio of 0. 36 bpp and 29. 86 dB in PSNR,its encoding time is 64 seconds and its decoding time is 109 seconds.

关 键 词：8邻域 分形块编码 定义域块 值域块 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]